Search

Summary: The retina extracts chromatic information present in an animal’s environment. How this information is processed in the retina is not well understood. In the mouse, chromatic information is not collected equally throughout the retina. Green and UV signals are primarily detected in the dorsal and ventral retina, respectively. However, at the output of the retina, chromatic tuning is more mixed throughout the retina. This suggests that lateral processing by inhibitory amacrine cells shapes chromatic information at the retinal output. We systematically surveyed the chromatic responses of dendritic processes of the 40+ GABAergic amacrine cell types. We identified 25 functional types with distinct chromatic and achromatic properties. We used pharmacology and a biologically inspired deep learning model to explore how inhibition and excitation shape the properties of functional types. Our data suggest that amacrine cells balance the biased spectral tuning of excitation, thereby supporting diversity of chromatic information at the retinal output.

The retina transforms patterns of light into visual feature representations supporting behaviour. These representations are distributed across various types of retinal ganglion cells (RGCs), whose spatial and temporal tuning properties have been studied extensively in many model organisms, including the mouse. However, it has been difficult to link the potentially nonlinear retinal transformations of natural visual inputs to specific ethological purposes. Here, we discover a nonlinear selectivity to chromatic contrast in an RGC type that allows the detection of changes in visual context. We trained a convolutional neural network (CNN) model on large-scale functional recordings of RGC responses to natural mouse movies, and then used this model to search in silico for stimuli that maximally excite distinct types of RGCs. This procedure predicted centre colour opponency in transient suppressed-by-contrast (tSbC) RGCs, a cell type whose function is being debated. We confirmed experimentally that these cells indeed responded very selectively to Green-OFF, UV-ON contrasts. This type of chromatic contrast was characteristic of transitions from ground to sky in the visual scene, as might be elicited by head or eye movements across the horizon. Because tSbC cells performed best among all RGC types at reliably detecting these transitions, we suggest a role for this RGC type in providing contextual information (i.e. sky or ground) necessary for the selection of appropriate behavioural responses to other stimuli, such as looming objects. Our work showcases how a combination of experiments with natural stimuli and computational modelling allows discovering novel types of stimulus selectivity and identifying their potential ethological relevance.

Neural population responses in sensory systems are driven by external physical stimuli. This stimulus-response relationship is typically characterized by receptive fields, which have been estimated by neural system identification approaches. Such models usually require a large amount of training data, yet, the recording time for animal experiments is limited, giving rise to epistemic uncertainty for the learned neural transfer functions. While deep neural network models have demonstrated excellent power on neural prediction, they usually do not provide the uncertainty of the resulting neural representations and derived statistics, such as most exciting inputs (MEIs), from in silico experiments. Here, we present a Bayesian system identification approach to predict neural responses to visual stimuli, and explore whether explicitly modeling network weight variability can be beneficial for identifying neural response properties. To this end, we use variational inference to estimate the posterior distribution of each model weight given the training data. Tests with different neural datasets demonstrate that this method can achieve higher or comparable performance on neural prediction, with a much higher data efficiency compared to Monte Carlo dropout methods and traditional models using point estimates of the model parameters. At the same time, our variational method provides us with an effectively infinite ensemble, avoiding the idiosyncrasy of any single model, to generate MEIs. This allows us to estimate the uncertainty of stimulus-response function, which we have found to be negatively correlated with the predictive performance at model level and may serve to evaluate models. Furthermore, our approach enables us to identify response properties with credible intervals and to determine whether the inferred features are meaningful by performing statistical tests on MEIs. Finally, in silico experiments show that our model generates stimuli driving neuronal activity significantly better than traditional models in the limited-data regime.

In the vertebrate retina, several dozens of parallel channels relay information about the visual world to the brain. These channels are represented by the different types of retinal ganglion cells (RGCs), whose responses are rendered selective for distinct sets of visual features by various mechanisms. These mechanisms can be roughly grouped into synaptic interactions and cell-intrinsic mechanisms, with the latter including dendritic morphology as well as ion channel complement and distribution. Here, we investigate how strongly ion channel complement can shape RGC output by comparing two mouse RGC types, the well-described ON alpha cell and a little-studied ON cell that is EGFP-labelled in the Igfbp5 mouse line and displays an unusual selectivity for stimuli with high contrast. Using patch-clamp recordings and computational modelling, we show that a higher activation threshold and a pronounced slow inactivation of the voltage-gated Na+ channels contribute to the distinct contrast tuning and transient responses in ON Igfbp5 RGCs, respectively. In contrast, such a mechanism could not be observed in ON alpha cells. This study provides an example for the powerful role that the last stage of retinal processing can play in shaping RGC responses.

Motion vision is critical for survival. Here the authors show that motion detection occurs already in bipolar cells of the mouse retina, which may contribute to motion processing throughout the visual system.

Neural system identification aims at learning the response function of neurons to arbitrary stimuli using experimentally recorded data, but typically does not leverage normative principles such as efficient coding of natural environments. Visual systems, however, have evolved to efficiently process input from the natural environment. Here, we present a normative network regularization for system identification models by incorporating, as a regularizer, the efficient coding hypothesis, which states that neural response properties of sensory representations are strongly shaped by the need to preserve most of the stimulus information with limited resources. Using this approach, we explored if a system identification model can be improved by sharing its convolutional filters with those of an autoencoder which aims to efficiently encode natural stimuli. To this end, we built a hybrid model to predict the responses of retinal neurons to noise stimuli. This approach did not only yield a higher performance than the "stand-alone" system identification model, it also produced more biologically plausible filters, meaning that they more closely resembled neural representation in early visual systems. We found these results applied to retinal responses to different artificial stimuli and across model architectures. Moreover, our normatively regularized model performed particularly well in predicting responses of direction-of-motion sensitive retinal neurons. The benefit of natural scene statistics became marginal, however, for predicting the responses to natural movies. In summary, our results indicate that efficiently encoding environmental inputs can improve system identification models, at least for noise stimuli, and point to the benefit of probing the visual system with naturalistic stimuli.

Since its invention, super-resolution microscopy has become a popular tool for advanced imaging of biological structures, allowing visualisation of subcellular structures at a spatial scale below the diffraction limit. Thus, it is not surprising that recently, different super-resolution techniques are being applied in neuroscience, e.g. to resolve the clustering of neurotransmitter receptors and protein complex composition in presynaptic terminals. Still, the vast majority of these experiments were carried out either in cell cultures or very thin tissue sections, while there are only a few examples of super-resolution imaging in deeper layers (30 - 50 µm) of biological samples. In that context, the mammalian whole-mount retina has rarely been studied with super-resolution microscopy. Here, we aimed at establishing a stimulated-emission-depletion (STED) microscopy protocol for imaging whole-mount retina. To this end, we developed sample preparation including horizontal slicing of retinal tissue, an immunolabeling protocol with STED-compatible fluorophores and optimised the image acquisition settings. We labelled subcellular structures in somata, dendrites, and axons of retinal ganglion cells in the inner mouse retina. By measuring the full width at half maximum of the thinnest filamentous structures in our preparation, we achieved a resolution enhancement of two or higher compared to conventional confocal images. When combined with horizontal slicing of the retina, these settings allowed visualisation of putative GABAergic horizontal cell synapses in the outer retina. Taken together, we successfully established a STED protocol for reliable super-resolution imaging in the whole-mount mouse retina at depths between 30 and 50 µm, which enables investigating, for instance, protein complex composition and cytoskeletal ultrastructure at retinal synapses in health and disease.

Abstract Hereditary degeneration of photoreceptors has been linked to over-activation of Ca2+-permeable channels, excessive Ca2+-influx, and downstream activation of Ca2+-dependent calpain-type proteases. Unfortunately, after more than 20 years of pertinent research, unequivocal evidence proving significant and reproducible photoreceptor protection with Ca2+-channel blockers is still lacking. Here, we show that both D- and L-cis enantiomers of the anti-hypertensive drug diltiazem were very effective at blocking photoreceptor Ca2+-influx, most probably by blocking the pore of Ca2+-permeable channels. Yet, unexpectedly, this block neither reduced the activity of calpain-type proteases, nor did it result in photoreceptor protection. Remarkably, application of the L-cis enantiomer of diltiazem even led to a strong increase in photoreceptor cell death. These findings shed doubt on the previously proposed links between Ca2+ and retinal degeneration and are highly relevant for future therapy development as they may serve to refocus research efforts towards alternative, Ca2+-independent degenerative mechanisms.

Mice are able to discriminate colors, at least in the upper visual field. Here, the authors provide a comprehensive characterization of retinal circuits underlying this behavior.

Neurons compute by integrating synaptic inputs across their dendritic arbor. Here, the authors show that distinct cell-types of mouse retinal ganglion cells that receive similar excitatory inputs have different biophysical mechanisms of input integration to generate their unique response tuning.

While multicompartment models have long been used to study the biophysics of neurons, it is still challenging to infer the parameters of such models from data including uncertainty estimates. Here, we performed Bayesian inference for the parameters of detailed neuron models of a photoreceptor and an OFF- and an ON-cone bipolar cell from the mouse retina based on two-photon imaging data. We obtained multivariate posterior distributions specifying plausible parameter ranges consistent with the data and allowing to identify parameters poorly constrained by the data. To demonstrate the potential of such mechanistic data-driven neuron models, we created a simulation environment for external electrical stimulation of the retina and optimized stimulus waveforms to target OFF- and ON-cone bipolar cells, a current major problem of retinal neuroprosthetics.

[This corrects the article DOI: 10.1371/journal.pcbi.1007205.].

Visual neuroscientists require accurate control of visual stimulation. However, few stimulator solutions simultaneously offer high spatio-temporal resolution and free control over the spectra of the light sources, because they rely on off-the-shelf technology developed for human trichromatic vision. Importantly, consumer displays fail to drive UV-shifted short wavelength-sensitive photoreceptors, which strongly contribute to visual behaviour in many animals, including mice, zebrafish and fruit flies. Moreover, many non-mammalian species feature more than three spectral photoreceptor types. Here, we present a flexible, spatial visual stimulator with up to six arbitrary spectrum chromatic channels. It combines a standard digital light processing engine with open source hard- and software that can be easily adapted to the experimentalist’s needs. We demonstrate the capability of this general visual stimulator experimentally in the in vitro mouse retinal whole-mount and the in vivo zebrafish. With this work, we intend to start a community effort of sharing and developing a common stimulator design for vision research.

Variability, stochastic or otherwise, is a central feature of neural activity. Yet the means by which estimates of variation and uncertainty are derived from noisy observations of neural activity is often heuristic, with more weight given to numerical convenience than statistical rigour. For two-photon imaging data, composed of fundamentally probabilistic streams of photon detections, the problem is particularly acute. Here, we present a statistical pipeline for the inference and analysis of neural activity using Gaussian Process regression, applied to two-photon recordings of light-driven activity in ex vivo mouse retina. We demonstrate the flexibility and extensibility of these models, considering cases with non-stationary statistics, driven by complex parametric stimuli, in signal discrimination, hierarchical clustering and other inference tasks. Sparse approximation methods allow these models to be fitted rapidly, permitting them to actively guide the design of light stimulation in the midst of ongoing two-photon experiments.

Understanding how neurons encode and compute information is fundamental to our study of the brain, but opportunities for hands-on experience with neurophysiological techniques on live neurons are scarce in science education. Here, we present Spikeling, an open source in silico implementation of a spiking neuron that costs £25 and mimics a wide range of neuronal behaviours for classroom education and public neuroscience outreach. Spikeling is based on an Arduino microcontroller running the computationally efficient Izhikevich model of a spiking neuron. The microcontroller is connected to input ports that simulate synaptic excitation or inhibition, to dials controlling current injection and noise levels, to a photodiode that makes Spikeling light sensitive, and to a light-emitting diode (LED) and speaker that allows spikes to be seen and heard. Output ports provide access to variables such as membrane potential for recording in experiments or digital signals that can be used to excite other connected Spikelings. These features allow for the intuitive exploration of the function of neurons and networks mimicking electrophysiological experiments. We also report our experience of using Spikeling as a teaching tool for undergraduate and graduate neuroscience education in Nigeria and the United Kingdom.

In recent years, two-photon calcium imaging has become a standard tool to probe the function of neural circuits and to study computations in neuronal populations. However, the acquired signal is only an indirect measurement of neural activity due to the comparatively slow dynamics of fluorescent calcium indicators. Different algorithms for estimating spike rates from noisy calcium measurements have been proposed in the past, but it is an open question how far performance can be improved. Here, we report the results of the spikefinder challenge, launched to catalyze the development of new spike rate inference algorithms through crowd-sourcing. We present ten of the submitted algorithms which show improved performance compared to previously evaluated methods. Interestingly, the top-performing algorithms are based on a wide range of principles from deep neural networks to generative models, yet provide highly correlated estimates of the neural activity. The competition shows that benchmark challenges can drive algorithmic developments in neuroscience.

In the mouse retina, three different types of photoreceptors provide input to 14 bipolar cell (BC) types. Classically, most BC types are thought to contact all cones within their dendritic field; ON-BCs would contact cones exclusively via so-called invaginating synapses, while OFF-BCs would form basal synapses. By mining publically available electron microscopy data, we discovered interesting violations of these rules of outer retinal connectivity: ON-BC type X contacted only ~20% of the cones in its dendritic field and made mostly atypical non-invaginating contacts. Types 5T, 5O and 8 also contacted fewer cones than expected. In addition, we found that rod BCs received input from cones, providing anatomical evidence that rod and cone pathways are interconnected in both directions. This suggests that the organization of the outer plexiform layer is more complex than classically thought.

[This corrects the article DOI: 10.1371/journal.pbio.1002086.].

The introduction of affordable, consumer-oriented 3-D printers is a milestone in the current "maker movement," which has been heralded as the next industrial revolution. Combined with free and open sharing of detailed design blueprints and accessible development tools, rapid prototypes of complex products can now be assembled in one's own garage--a game-changer reminiscent of the early days of personal computing. At the same time, 3-D printing has also allowed the scientific and engineering community to build the "little things" that help a lab get up and running much faster and easier than ever before.

Cell death in neurodegenerative diseases is often thought to be governed by apoptosis; however, an increasing body of evidence suggests the involvement of alternative cell death mechanisms in neuronal degeneration. We studied retinal neurodegeneration using 10 different animal models, covering all major groups of hereditary human blindness (rd1, rd2, rd10, Cngb1 KO, Rho KO, S334ter, P23H, Cnga3 KO, cpfl1, Rpe65 KO), by investigating metabolic processes relevant for different forms of cell death. We show that apoptosis plays only a minor role in the inherited forms of retinal neurodegeneration studied, where instead, a non-apoptotic degenerative mechanism common to all mutants is of major importance. Hallmark features of this pathway are activation of histone deacetylase, poly-ADP-ribose-polymerase, and calpain, as well as accumulation of cyclic guanosine monophosphate and poly-ADP-ribose. Our work thus demonstrates the prevalence of alternative cell death mechanisms in inherited retinal degeneration and provides a rational basis for the design of mutation-independent treatments.

Far from being a simple sensor, the retina actively participates in processing visual signals. One of the best understood aspects of this processing is the detection of motion direction. Direction-selective (DS) retinal circuits include several subtypes of ganglion cells (GCs) and inhibitory interneurons, such as starburst amacrine cells (SACs). Recent studies demonstrated a surprising complexity in the arrangement of synapses in the DS circuit, i.e. between SACs and DS ganglion cells. Thus, to fully understand retinal DS mechanisms, detailed knowledge of all synaptic elements involved, particularly the nature and localization of neurotransmitter receptors, is needed. Since inhibition from SACs onto DSGCs is crucial for generating retinal direction selectivity, we investigate here the nature of the GABA receptors mediating this interaction. We found that in the inner plexiform layer (IPL) of mouse and rabbit retina, GABA(A) receptor subunit α2 (GABA(A)R α2) aggregated in synaptic clusters along two bands overlapping the dendritic plexuses of both ON and OFF SACs. On distal dendrites of individually labeled SACs in rabbit, GABA(A)R α2 was aligned with the majority of varicosities, the cell's output structures, and found postsynaptically on DSGC dendrites, both in the ON and OFF portion of the IPL. In GABA(A)R α2 knock-out (KO) mice, light responses of retinal GCs recorded with two-photon calcium imaging revealed a significant impairment of DS responses compared to their wild-type littermates. We observed a dramatic drop in the proportion of cells exhibiting DS phenotype in both the ON and ON-OFF populations, which strongly supports our anatomical findings that α2-containing GABA(A)Rs are critical for mediating retinal DS inhibition. Our study reveals for the first time, to the best of our knowledge, the precise functional localization of a specific receptor subunit in the retinal DS circuit.

Detection of image motion direction begins in the retina, with starburst amacrine cells (SACs) playing a major role. SACs generate larger dendritic Ca(2+) signals when motion is from their somata towards their dendritic tips than for motion in the opposite direction. To study the mechanisms underlying the computation of direction selectivity (DS) in SAC dendrites, electrical responses to expanding and contracting circular wave visual stimuli were measured via somatic whole-cell recordings and quantified using Fourier analysis. Fundamental and, especially, harmonic frequency components were larger for expanding stimuli. This DS persists in the presence of GABA and glycine receptor antagonists, suggesting that inhibitory network interactions are not essential. The presence of harmonics indicates nonlinearity, which, as the relationship between harmonic amplitudes and holding potential indicates, is likely due to the activation of voltage-gated channels. [Ca(2+)] changes in SAC dendrites evoked by voltage steps and monitored by two-photon microscopy suggest that the distal dendrite is tonically depolarized relative to the soma, due in part to resting currents mediated by tonic glutamatergic synaptic input, and that high-voltage-activated Ca(2+) channels are active at rest. Supported by compartmental modeling, we conclude that dendritic DS in SACs can be computed by the dendrites themselves, relying on voltage-gated channels and a dendritic voltage gradient, which provides the spatial asymmetry necessary for direction discrimination.

Genetically encoded fluorescent calcium indicator proteins (FCIPs) are promising tools to study calcium dynamics in many activity-dependent molecular and cellular processes. Great hopes-for the measurement of population activity, in particular-have therefore been placed on calcium indicators derived from the green fluorescent protein and their expression in (selected) neuronal populations. Calcium transients can rise within milliseconds, making them suitable as reporters of fast neuronal activity. We here report the production of stable transgenic mouse lines with two different functional calcium indicators, inverse pericam and camgaroo-2, under the control of the tetracycline-inducible promoter. Using a variety of in vitro and in vivo assays, we find that stimuli known to increase intracellular calcium concentration (somatically triggered action potentials (APs) and synaptic and sensory stimulation) can cause substantial and rapid changes in FCIP fluorescence of inverse pericam and camgaroo-2.